Gene Selection for Cancer Classification Using DCA

The problem of gene selection for cancer classification is considered.A combined SVM-feature selection approach based on the smoothly clipped absolute deviation(SCAD)penalty is developed,minimizing directly the classifier performance.To solve the optimization problems,apply the DCA(difference of convex functions algorithms) which is a general framework for nonconvex continuous optimization.This leads to a successive linear programming algorithm with finite convergence.Preliminary computational experiments on different real data demonstrate that this method accomplishes the desired goal:Suppression of a large number of features with a small error of classification.